Disorders of communication are core deficits in autism. Children with ASD not only have problems with expressive language, they also show deficits in understanding, as well as early failures to orient to speech and name. This project aims to examine precursors of these deficits by invetsigating auditory preferences known to be present in typically developing infants and to support the growth of social engagment and language. We aim to test the following hypotheses: 1) older infants (12-24 mo.) with symptoms of ASD will show reduced preference, relative to typical and developmental^ delayed peers, for listening to speech under controlled experimental conditions, as opposed to nonsocial auditory stimuli with analogous acoustic properties;2) they will show reduced preference, relative to TD and DD peers, for speech with "motherese" prosody, as opposed to speech with neutral prosody;3) they will show reduced experience-based preferences for word forms with the predominant stress and phonotactic patterns of English, relative to TD and DD peers;4) listening preferences at 12-24 mo. (Time 1) will discriminate children who meet diagnostic criteria for ASD from those with TD and DD, when diagnosis is confirmed at 36-48 mo. (Time 2);and 5) these listening preferences will be correlated with concurrent degree of communicative and social competence at Time 1, and will predict language and social competence at Time 2. We request 5 years of support to examine these hypotheses in one year-olds and to follow the ASD and DD groups to 36-48 mo. For Hypotheses 1-3, the infant-controlled visual fixation paradigm, a method used for the last ten years in the study of infant speech perception, will be used to examine group differences in auditory preferences at Time 1. For Hypotheses 4-5, data from the auditory preference experiments will be used to predict diagnostic outcome categorically at Time 2, as well as to predict dimensional measures of social, cognitive, language, and adaptive development, both cross-sectionally (at Time 1), and longitudinally (at Time 2). Longer-term goals are to develop these paradigms, in conjunction with the visual measures studied in Projects 1 and 2, into infant screening methods for ASD, as the IACC Autism Research Matrix advocates. This information will also be used to generate hypotheses regarding treatment to optimize early auditory experiences, in efforts aimed at achieving the Matrix goal of increasing the prevalence of spoken langugage in ASD to 90%.